gender identity disorder and eating disorders: similarities and differences in terms of body...

ORIGINAL RESEARCH—INTERSEX AND GENDERIDENTITY DISORDERS

Gender Identity Disorder and Eating Disorders: Similarities andDifferences in Terms of Body Uneasiness

Elisa Bandini, MD,* Alessandra Daphne Fisher, MD,* Giovanni Castellini, MD, PhD,†

Carolina Lo Sauro, MD,† Lorenzo Lelli, MD,† Maria Cristina Meriggiola, MD,‡ Helen Casale, Psy D,*Laura Benni, MD,† Naika Ferruccio, MD,† Carlo Faravelli, MD,§ Davide Dettore, Psy D,¶

Mario Maggi, MD,* and Valdo Ricca, MD†

*Andrology Unit, Department of Clinical Physiopathology, Florence University School of Medicine, Firenze, Italy;†Psychiatric Unit, Department of Neuropsychiatric Sciences, Florence University School of Medicine, Firenze, Italy;‡Center for Reproductive Health, Department of Obstetrics and Gynecology, S. Orsola Hospital, University of Bologna,Bologna, Italy; §Department of Psychology, University of Florence, Firenze, Italy; ¶Department of Health Sciences,University of Florence, Firenze, Italy

DOI: 10.1111/jsm.12062

A B S T R A C T

Introduction. Subjects with gender identity disorder (GID) have been reported to be highly dissatisfied with theirbody, and it has been suggested that the body is their primary source of suffering.Aims. To evaluate quality and intensity of body uneasiness in GID subjects, comparing them with a sample of eatingdisorder patients and a control group. To detect similarities and differences between subgroups of GID subjects, onthe basis of genotypic sex and transitional stage.Methods. Fifty male-to-female (MtF) GID (25 without and 25 with genital reassignment surgery performed), 50female-to-male (FtM) GID (28 without and 22 with genital reassignment surgery performed), 88 eating disorder subjects(26 anorexia nervosa, 26 bulimia nervosa, and 36 binge eating disorder), and 107 healthy subjects were evaluated.Main Outcome Measures. Subjects were studied by means of the Structured Clinical Interview for Diagnostic andStatistical Manual of Mental Disorders, Fourth Edition, the Symptom Checklist (SCL-90), and the Body UneasinessTest (BUT).Results. GID and controls reported lower psychiatric comorbidity and lower SCL-90 General Severity Index (GSI)scores than eating disorder subjects. GID MtF without genital reassignment surgery showed the highest BUT values,whereas GID FtM without genital reassignment surgery and eating disorder subjects showed higher values comparedwith both GID MtF and FtM who underwent genital reassignment surgery and controls. Considering BUTsubscales, a different pattern of body uneasiness was observed in GID and eating disorder subjects. GID MtF andFtM without genital reassignment surgery showed the highest BUT GSI/SCL-90 GSI ratio compared with all theeating disorder groups.Conclusions. GID and eating disorders are characterized by a severe body uneasiness, which represents the core ofdistress in both conditions. Different dimensions of body uneasiness seem to be involved in GID subsamples,depending on reassignment stage and genotypic sex. In eating disorder subjects body uneasiness is primarily linkedto general psychopathology, whereas in GID such a relationship is lacking. Bandini E, Fisher AD, Castellini G, LoSauro C, Lelli L, Meriggiola MC, Casale H, Benni L, Ferruccio N, Faravelli C, Dettore D, Maggi M, andRicca V. Gender identity disorder and eating disorders: Similarities and differences in terms of bodyuneasiness. J Sex Med 2013;10:1012–1023.

Key Words. Body Uneasiness; Gender Identity Disorder; Eating Disorder; Psychopathology

Elisa Bandini and Alessandra Daphne Fisher equally participated in this study.

1012

J Sex Med 2013;10:1012–1023 © 2013 International Society for Sexual Medicine

Introduction

G ender identity disorder (GID) is character-ized by a strong and persistent identification

with the opposite sex, discomfort with one’s ownsex, and a sense of inappropriateness in the genderrole of that sex [1]. GID subjects live in a cognitivestate where their physical body is in contrast withtheir self-perceived identity [2], and this experi-ence is a source of deep and chronic suffering [3].As far as psychopathology is concerned, most dataconfirm that GID subjects lack severe primarypsychopathology [4–7], although some studieshave reported a relatively high prevalence ofpsychiatric disorders among GID patients [8,9].Nevertheless, it has been suggested that psychopa-thology may be the consequence of difficulties incoping with GID [6], social stigma [10–12], orrejection by family and friends [13], rather thanfrom a primary psychiatric illness. Indeed, hor-monal and surgical sex reassignment have beenreported to improve quality of life and psychologi-cal well-being [14,15]. Moreover, some studiesreported that GID subjects are more dissatisfiedwith their body than subjects without GID, evenwith regard to nonsexual body parts and aspects[16,17], suggesting that the body is the primarysource of suffering in these persons. The centralityof the body in gender dysphoria is also demon-strated by the fact that its successful treatment isalso capable of relieving general psychopathologyand distress [18].

As far as eating disorders are concerned, a maindistressing factor is the body, as well. In fact,according to the cognitive model of the mainte-nance of eating disorders [19] the “core psychopa-thology” of these disorders is an overconcernabout body shape and body weight, and self-worthis judged largely or even exclusively in termsof satisfaction–unsatisfaction with them. Bodyuneasiness and dissatisfaction are relevant modera-tors of current treatments [20,21] and, even aftersuccessful management of abnormal eating behav-ior, their persistence is a predictor of relapse [22].Unlike GID, eating disorders show relevant psy-chiatric comorbidity [23], as about 80% of patientsare diagnosed with another psychiatric disorder atsome time in their life [24]. In particular, moodand anxiety disorders, substance abuse, as well asobsessive–compulsive and borderline personalitytraits affect a large proportion of eating disorderpatients and can persist even after recovery [25].

Until now, only one empirical study hadexplored the degree of eating and body image dis-

turbance, self-esteem, and depression in a sampleof GID subjects compared with an eating disordersclinical sample and a control group [17]. However,this study had some methodological limitations.First, the GID subjects were self-identified. More-over, the clinical subtypes of eating disorders werenot taken in account, and sexual orientation, whichis known to have an effect on body image, wasnot assessed, as well as general psychopathology.Finally, the GID sample was not divided intosubsamples, according to transitional stages.

In conclusion, similarities and differencesbetween GID and eating disorder patients in termsof intensity and quality of body uneasiness, andgeneral psychopathology, are far from having beenelucidated.

Aims

The aims of the present study are

• To compare a sample of subjects with GID,a sample of subjects with eating disorders,and a control group in terms of quality andintensity of body uneasiness and generalpsychopathology.

• To evaluate similarities and differences betweensubgroups of GID subjects, identified on thebasis of genotypic sex and transitional stage,along with these psychopathological features.

Methods and Main Outcome Measures

The study was designed as a cross-sectional survey,and it was planned and performed at the Centerfor Assistance to GID of the University ofFlorence, in a dedicated Center for GID atBologna, and at the Outpatient Clinic for EatingDisorders of the University of Florence.

Participants were recruited from consecutivereferrals by family doctors and other clinicians. Allthe diagnostic procedures and the psychometrictests are part of the clinical and psychodiagnosticroutine procedure for GID and eating disorders.The study protocol was approved by the localEthics Committee. Before the collection of data,during the first routine visit, the procedures of thestudy were fully explained. After that, the patientswere asked to provide their written informedconsent to the participation in the present study.The study was carried out in accordance with theethical standards of the responsible institutionalcommittee.

GID and Eating Disorders: The Role of the Body 1013

J Sex Med 2013;10:1012–1023

ParticipantsGID SubjectsAll the subjects referring for the first time to thetwo GID clinics were enrolled in the study, pro-vided they met the following inclusion criteria:

• Age range between 18 and 60 years• Diagnosis of GID, based on formal psychiatric

classification criteria [1,10] and performedthrough several sessions with two differentmental health professionals specialized in GID

The exclusion criteria were as follows:

• Genital reassignment surgery (GRS) performedin the last 24 months or GRS performedwithout 12 months of prior cross sex hormonaltherapy

• Taking or have taken cross sex hormones duringtheir life when GRS had not been performed

• Illiteracy• Mental retardation• Presence of any actual or lifetime eating disor-

der diagnosis

One hundred fourteen subjects were excludedfrom the initial sample (N = 214) because of thefollowing reasons: taking or have taken hormonesduring their life (N = 95), GRS performed in thelast 24 months (N = 9), GRS performed without12 months of prior hormonal therapy (N = 1), illit-eracy (N = 2), mental retardation (N = 3), actual orprevious eating disorder diagnosis (N = 4).

The selected sample (N = 100, 59 from Flo-rence, and 41 from Bologna) was divided into fourgroups according to their genotypic sex and GRSperformed. In particular, male-to-female (MtFs)and female-to-male subjects (FtMs) were equallyrepresented (50 subjects in each group). In addi-tion, 50% of MtFs and 56% of FtMs underwentGRS at least 24 months before their enrolment inthe study, performed after 12 months of prior crosssex hormonal therapy.

Eating Disorder SubjectsAll the subjects referring for the first time to theClinic for Eating Disorders of the University ofFlorence, between July 2010 and June 2011, wereenrolled in the study, provided they met thefollowing inclusion criteria:

• Age range between 18 and 60 years• Diagnosis of eating disorder, according to Diag-

nostic and Statistical Manual of Mental Disor-ders, Fourth Edition (DSM-IV) [1] criteria,performed by means of the Structured ClinicalInterview for DSM-IV (SCID I) [26]

The exclusion criteria were as follows:

• Illiteracy• Mental retardation• Presence of any actual or lifetime GID diagnosis

The initial sample (N = 94) was divided into threegroups according to different eating disorderDSM-IV diagnoses. In particular 27 subjects metdiagnostic criteria for anorexia nervosa (AN), 28for bulimia nervosa (BN), and 39 for binge eatingdisorder (BED). Of the 94 eating disorder sub-jects initially enrolled in the study, four subjectswere excluded for illiteracy and two for mentalretardation.

Control GroupThis group was enrolled according to the follow-ing procedure: each control was extracted from thealphabetical computerized list of clients of ageneral practitioner and was selected as the firstone fulfilling the inclusion criteria and willing toparticipate. In the case of refusal, the next one onthe list fitting the matching criteria was asked toparticipate.

The control inclusion criteria were as follows:

• Age range between 18 and 60 years• Body mass index (BMI) ranging between 18 and

24.9 kg/m2

The exclusion criteria were the same as those usedfor the patient groups, plus the presence of anycurrent and lifetime GID or eating disorder diag-nosis, according to DSM-IV criteria [1].

Thirty-one subjects were excluded from theinitial list of controls because of the followingreasons: refusal to give their informed consent(N = 22), illiteracy (N = 6), and mental retardation(N = 3). The selected sample was made up of 107subjects.

MeasuresThe sociodemographic data, as well as the anthro-pometric measures, were collected at the begin-ning of the first visit, through a face-to-faceinterview, by mental health specialists expert inthese fields (VR, EB, GC, HC). Subjects reportedtheir age, relationship status, morbidities, andreferred sexual orientation.

Anthropometric measurements were madeusing standard calibrated instruments during theroutine psychiatric visits.

The presence of Axis I psychiatric comorbiditywas evaluated by a mental health specialist usingthe SCID I [26].

1014 Bandini et al.

J Sex Med 2013;10:1012–1023

Furthermore, the patients were evaluated bymeans of the Italian version of the SymptomChecklist (SCL-90-R) [27,28], a psychometricinstrument devoted to the identification of thepsychopathological distress, during the previousweek. The 90 items of the questionnaire, rated ona five-point Likert-type scale (from 0 to 4), aregrouped together into nine domains (somatiza-tion, obsessive–compulsive thoughts, interper-sonal sensitivity, depression, anxiety, hostility,phobic anxiety, paranoid conceiving, and psychoticbehavior), and a General Severity Index (GSI),indicating the overall psychological distress. Theinternal consistencies (coefficient alpha) concern-ing the nine primary symptom scales are rankedbetween a low of 0.77 for psychoticism and a highof 0.90 for depression [29]. Test–retest reliabilityhas been reported at 0.80 to 0.90 with a timeinterval of 1 week [28].

For a specific assessment of body uneasinessand dissatisfaction, the Body Uneasiness Test(BUT) [30] was applied. The BUT is a self-ratingscale that simultaneously explores various areas ofbody-related psychopathology. Subjects are askedto rate 34 different body experiences (BUT A)and 37 body parts (BUT B), on a six-point Likert-type scale (from never to always), indicating: (i)dissatisfaction regarding the body and its weight(e.g., “I’m terrified of gaining weight,” “Myphysical appearance is unsatisfying compared tomy ideal body image”); (ii) avoiding and compul-sive self-monitoring behavior (e.g., “When I getundressed I avoid looking,” “I spend a lot of timein front of the mirror,” “I fear that my appearancesuddenly change”); (iii) experience of separationand foreignness regarding the body (e.g., “WhenI look at myself in the mirror I feel a sense ofanxiety and alienation”); and (iv) specific worriesabout certain body parts, characteristics or func-tions (e.g., skin, mouth, breasts, knees, mous-taches, hair, smell, noise, sweating, flushing).Higher scores mean greater body uneasiness.BUT scores were analyzed considering the totalscore of the test (Global Severity Index, BUTGSI) and several subscales (weight phobia, bodyimage concerns, avoidance, compulsive self-monitoring, and depersonalization).

BUT has been recently validated in a large non-clinical sample of individuals from adolescence toold age and in a large clinical sample of individualssuffering from eating disorders, showing goodpsychometric properties [31]. Internal consistencywas satisfactory in the validation study (Cron-bach’s value: weight phobia, 0.84; body image

concerns, 0.90; avoidance, 0.79; compulsive self-monitoring, 0.82; depersonalization, 0.85), as wellas when calculated in our sample (Cronbach’svalue: weight phobia, 0.91; body image concerns,0.94; avoidance, 0.88; compulsive self-monitoring,0.86; depersonalization, 0.90). The test–retest cor-relation coefficients were highly significant (BUTglobal severity index, 0.90) [31]. Finally, explor-atory and confirmatory analyses found the samestructural model for BUT A in normal-weight,non-eating-disordered subjects [30], as well as inpatients with obesity [32].

Statistical AnalysisSeven different groups were considered for theanalyses: healthy control subjects, MtFs with GRS(wGRS) and without GRS (w/oGRS) having beenperformed, FtMs wGRS and w/oGRS, AN, BN,and BED patients. For the assessment of between-group differences, chi-square and univariate analy-sis of variance (anova, with Tukey post hoc test)were applied for categorical and continuous vari-ables, respectively. The BUT GSI/SCL-90 ratiowas also calculated, in order to evaluate the inten-sity of body distress independently from the levelof psychopathology. Furthermore, in order to haveage- and BMI-adjusted analyses for between groupcomparisons, linear regression analyses were per-formed with SCL-90 GSI, BUT GSI, and BUTGSI/SCL-90 GSI ratio as dependent variables.Age and BMI were entered into the models ascovariates and groups were entered as dummyvariables (e.g., controls = 0; MtF wGRS = 1). Allanalyses were performed using SPSS for Windows15.0 (SPSS Inc., Chicago, IL, USA).

Results

General characteristics of the sample are reportedin Table 1. Comparison groups did not show sig-nificant differences in terms of socioeconomicstatus. GID wGRS and BED patients were olderthan the other groups. BMI was lower in AN andBN, and higher in BED than in the rest of thesample (all P < 0.05). No significant differenceswere found between male and female control sub-jects in terms of age and BMI. MtFs and FtMsw/oGRS showed a lower rate of plastic surgerycompared with GID wGRS (both P < 0.001).FtMs wGRS reported a higher rate of mamma-plasty and plastic surgery, compared with MtFswGRS (both P < 0.05) and FtMs w/oGRS (both


J Sex Med 2013;10:1012–1023

Tab

le1

The

tabl

ere

port

sth

ege

nera

lcha

ract

eris

tics

ofth

esa

mpl

e

Con

trol

s(N

=10

7,F

:46

.7%

)

MtF

GID

(N=

50)

FtM

GID

(N=

50)

Eat

ing

diso

rder

s(N

=88

)

w/o

GR

S(N

=25

)w

GR

S(N

=25

)w

/oG

RS

(N=

28)

wG

RS

(N=

22)

AN

(N=

26,

F:

76.9

%)

BN

(N=

26,

F:

96.2

%)

BE

D(N

=36

,F

:41

.7%

)

Age

(yea

rs)

33.6

7(�

9.24

)30

.44

(�10

.5)

42.9

2(�

11.1

9)27

.96

(�6.

9)38

.45

(�5.

92)

28(�

10.6

1)30

.85

(�9.

82)

39.9

2(�

14.2

8)B

ody

mas

sin

dex

(Kg/

m2 )

22.0

4(�

2.17

)21

.74

(�3.

03)

23.5

4(�

3.59

)25

.33

(�9.

68)

25.2

9(�

2.97

)16

.3(�

2.51

)20

.55

(�4.

24)

38.5

5(�

5.64

)S

tabl

ere

latio

nshi

p68

.2%

(N=

73)

32%

(N=

8)36

%(N

=9)

50%

(N=

14)

45.5

%(N

=10

)42

.3%

(N=

11)

80.8

%(N

=21

)52

.8%

(N=

19)

Sex

ualo

rient

atio

nTo

war

dsa

me

geno

typi

cse

x4.

7%(N

=5)

80%

(N=

20)

92%

(N=

23)

82.1

%(N

=23

)92

.3%

(N=

24)

0%(N

=0)

13.9

%(N

=5)

13.9

%(N

=5)

Tow

ard

oppo

site

geno

typi

cse

x90

.6%

(N=

97)

8%(N

=2)

4%(N

=1)

3.6%

(N=

1)7.

7%(N

=2)

100%

(N=

26)

83.3

%(N

=30

)83

.3%

(N=

30)

Tow

ard

both

sexe

s4.

7%(N

=5)

8%(N

=2)

4%(N

=1)

10.7

%(N

=3)

0%(N

=0)

0%(N

=0)

2.8%

(N=

1)2.

8%(N

=1)

Non

e0%

(N=

0)4%

(N=

1)0%

(N=

0)3.

6%(N

=1)

0%(N

=0)

0%(N

=0)

0%(N

=0)

0%(N

=0)

Pla

stic

surg

ery

0.9%

(N=

1)20

%(N

=5)

68%

(N=

42)

7.1%

(N=

4)86

.4%

(N=

42)

0%(N

=0)

0%(N

=0)

0%(N

=0)

Fac

ial

0%(N

=0)

8.0%

(N=

2)40

%(N

=10

)7.

1(N

=2)

18.2

(N=

4)0%

(N=

0)0%

(N=

0)0%

(N=

0)M

amm

apla

sty

0%(N

=0)

4%(N

=1)

60%

(N=

15)

0%(N

=0)

86.4

(N=

19)

0%(N

=0)

0%(N

=0)

0%(N

=0)

Oth

ers

0.9%

(N=

1)8%

(N=

2)68

%(N

=17

)7.

1%(N

=2)

86.4

%(N

=19

)0%

(N=

0)0%

(N=

0)0%

(N=

0)P

sych

iatr

icco

mor

bidi

ty12

.1%

(N=

13)

24%

(N=

6)16

%(N

=4)

14.3

%(N

=4)

13.6

%(N

=3)

42.3

%(N

=11

)46

.2%

(N=

12)

50.0

%(N

=18

)S

ubst

ance

abus

e0%

(N=

0)0%

(N=

0)8%

(N=

2)0%

(N=

0)0%

(N=

0)0%

(N=

0)0%

(N=

0)0%

(N=

0)U

nipo

lar

depr

essi

on1.

9%(N

=2)

8%(N

=2)

8%(N

=2)

7.1%

(2)

4.5%

(N=

1)26

.9%

(N=

7)23

.1%

(N=

6)27

.8%

(N=

10)

Bip

olar

diso

rder

0%(N

=0)

4%(N

=1)

0%(N

=0)

0%(N

=0)

0%(N

=0)

7.7%

(N=

2)7.

7%(N

=2)

11.1

%(N

=4)

Adj

ustm

ent

diso

rder

3.7%

(N=

4)16

%(N

=4)

0%(N

=0)

10.7

%(N

=3)

4.5%

(N=

1)0%

(N=

0)0%

(N=

0)0%

(N=

0)O

bses

sive

–com

puls

ive

diso

rder

3.7%

(N=

4)0%

(N=

0)0%

(N=

0)0%

(N=

0)0%

(N=

0)3.

8%(N

=1)

0%(N

=0)

8.3%

(N=

3)P

anic

diso

rder

7.5%

(N=

8)0%

(N=

0)4%

(N=

1)0%

(N=

0)4.

5%(N

=1)

3.8%

(N=

1)0%

(N=

0)2.

8%(N

=1)

Oth

ers

2.8%

(N=

3)8%

(N=

2)0%

(N=

0)3.

6%(N

=1)

0%(N

=0)

0%(N

=0)

15.4

%(N

=4)

2.8%

(N=

1)

Dat

aar

eex

pres

sed

asm

ean

�st

anda

rdde

viat

ion

whe

nno

rmal

lydi

strib

uted

,an

das

perc

enta

ges

and

abso

lute

num

ber,

repo

rted

inbr

acke

ts,

whe

nca

tego

rical

.G

ID=

gend

erid

entit

ydi

sord

er;

MtF

=m

ale

tofe

mal

e;F

tM=

fem

ale

tom

ale;

w/o

GR

S=

with

out

geni

tal

reas

sign

men

tsu

rger

y;w

GR

S=

with

geni

tal

reas

sign

men

tsu

rger

y;A

N=

anor

exia

nerv

osa;

BN

=bu

limia

nerv

osa;

BE

D=

bing

eea

ting

diso

rder

1016 Bandini et al.

J Sex Med 2013;10:1012–1023

P < 0.01). When adjusted for age, only differencesbetween FtMs wGRS and FtM w/oGRS inmammaplasty and between FtM wGRS and MtFwGRS in plastic surgery retained significance (allP < 0.05).

All GID subjects reported a higher frequencyof sexual orientation toward the same genotypicsex than controls and all eating disorder patients(all P < 0.001). Moreover, no differences wereobserved among controls and eating disorderpatients, with the exception of BED, who moreoften showed sexual orientation toward the samegenotypic sex than BN (c2 = 3.928, P = 0.047).Controls and BN patients more frequentlyreported a stable relationship than GID and ANand BED patients (all P < 0.05).

Psychiatric comorbidity was more frequent inall eating disorder groups than in controls (allP < 0.01), and than in GID subjects (all P < 0.01),with the exception of MtFs w/oGRS, who did notdiffer from AN and BN subjects. Bipolar disorderwas more prevalent in MtFs w/oGRS (c2 = 4.313,P = 0.038) and in all eating disorder patients (allP < 0.01) than in controls, and unipolar depressionwas more prevalent in all eating disorder patientsthan in controls (all P < 0.01). Adjustment disorderwas more prevalent in MtFs w/oGRS than incontrols (c2 = 5.352, P = 0.021), MtFs wGRS(c2 = 4.348, P = 0.037) and all eating disorder (allP < 0.05).

Figures 1–3 (upper section) show the distribu-tion of psychopathological variables (SCL-90 andBUT) according to the different groups. Univari-ate between-group comparisons revealed thatAN, BN, and BED patients showed higherSCL-90 GSI scores (without differences amongthem), compared with GID groups, which, inturn, did not show significant differences withcontrols (Figure 1, upper section). No significantdifference was found in terms of SCL-90 GSIscores between male and female controls, and thesame pattern of results was obtained when com-paring groups of patients with male and femalecontrols separately. Linear regression analysisrevealed that when adjusting for age and BMI,MtFs wGRS, MtFs w/oGRS, and FtMs wGRSdid not show significant differences when com-pared with BED subjects, whereas MtFs w/oGRSshowed higher values compared with controls(Figure 1, lower section). The same pattern ofdifferences was observed when considering allthe SCL-90 subscales (somatization, obsessive–compulsive thoughts, interpersonal sensitivity,depression, anxiety, hostility, phobic anxiety,

paranoid conceiving, and psychotic behavior; allP < 0.05).

As far as BUT scores are concerned, MtFsand FtMs w/oGRS showed the highest BUTGSI values, according to univariate analyses.In fact, these subjects, along with AN, BN, andBED patients, showed higher scores comparedwith GID wGRS and controls (Figure 2, uppersection). No significant difference in terms of BUTGSI was detected between male and female con-trols, and the same pattern of results was obtainedwhen comparing groups of patients with male andfemale controls separately. Linear regression analy-ses revealed that after adjusting for age and BMI, nosignificant difference was observed between GIDw/oGRS and BN subjects. In the adjusted model,the aforementioned differences retained signifi-cance (see Figure 2, lower section).

Figure 3 shows BUT subscale scores. Allgroups, with the exception of GID wGRS, showedhigher BUT weight phobia scores compared withcontrols, with MtF w/oGRS and BN showingthe highest. GID w/oGRS and eating disorderpatients, but not the other groups, showed higherBUT body image concerns scores compared withcontrols. All groups, with the exception of MtFswGRS, showed higher BUT avoidance scorescompared with controls, with MtFs w/oGRS andBED patients showing the highest. All groups,with the exception of FtMs wGRS and w/oGRS,showed higher BUT compulsive self-monitoringscores compared with controls, with MtFs wGRS,AN and BN patients reporting the highest. All thepatient groups, with the exception of GID wGRS,showed higher BUT depersonalization scorescompared with controls, with GID w/oGRS andBED subjects reporting the highest.

Considering the BUT/SCL-90 GSI ratio(Figure 4), GID w/oGRS showed higher valuesthan all the other groups, including eating disor-ders, which resulted similar to controls, even whenadjusting for age and BMI (Figure 4, lower section).

Discussion

To our knowledge, this is the first study that hasevaluated the body uneasiness in GID subjects,taking into account GID subsamples and compar-ing them with eating disorder patients. The mainresults of the present study are as follows:

• GID subjects w/oGRS generally showed highlevel of body uneasiness, which is independentfrom general psychopathology and higher whencompared with eating disorder patients.


J Sex Med 2013;10:1012–1023

• GID subjects wGRS showed lower body dissat-isfaction compared with those w/oGRS.

• In GID, general psychopathology was lowerthan in eating disorders and essentially similarto controls.

• MtFs and FtMs did not differ in terms of bodyuneasiness, although different dimensions seemto be involved in the two groups.

A higher level of body uneasiness was observed inGID w/oGRS compared with eating disorderpatients. Moreover, the ratio of BUT/SCL-90—arough index of body uneasiness adjusted forgeneral psychopathology—was 3–4 times higherin GID subjects w/oGRS than in eating disorders

and controls, respectively. This indicates that thebody uneasiness was almost independent from anunderlying psychiatric comorbidity and suggestsits pivotal role in the psychopathology of thesepatients.

Considering the differences between GID sub-jects with and without GRS, we found that in thefirst group, body uneasiness was dramaticallyreduced, as most of them reached control values.This result is in line with different findings sug-gesting the positive effect of GRS on body dissat-isfaction in GID subjects, due to a reduction inthe discrepancy between biological and desiredsex [33–35]. Moreover, no difference in bodyuneasiness was observed between MtFs and FtMs,

Figure 1 SCL-90 Global Severity Index: comparisons among groups. Statistics—Values reported in the table are b valuesof Linear Regression Analyses, performed for between groups comparisons wih SCL-90 GSI as dependent variable, andentering age, and BMI as covariates, and groups as dummy variables (e.g., controls = 0; MtF w GRS = 1). *P < 0.05;**P < 0.01; ***P < 0.001. CL-90 GSI = Sympton Checklist Global Severity Index; MtF = male to female; FtM = female to male;GRS = genital reassignment surgery (w = performed; w/o = not performed); AN = anorexia nervosa; BN = bulimia nervosa;BED = binge eating disorder

1018 Bandini et al.

J Sex Med 2013;10:1012–1023

as previously reported [17]. As far as BUT sub-scales scores are concerned, the higher value ofbody image concerns in GID w/oGRS subjectssupports the positive effects of reassignment onsuch preoccupations.

GID patients w/oGRS displayed higher levelsof depersonalization. This result confirms twoprevious studies suggesting that the depersonaliza-tion experience is less intense in patients who haveundergone GRS [36], and that MtFs wGRS didnot differ from female and male controls in termsof depersonalization and body satisfaction [37].The higher avoidance observed in the GID samplew/oGRS suggests that GID is highly distressing.Considering that GRS may help normalize/reduce

avoidance to the same level of controls in MtFsbut not in FtMs, it could be speculated thatthis depends on different outcomes associatedwith reassignment surgery in the two genders.However, in our sample only few FtMs subjectshad performed reconstructive genital surgery.

Weight phobia and compulsive body self-monitoring scores were higher in MtFs than FtMs.A previous research reported that MtFs displayedmore weight concerns than male and female con-trols [17]. In the general population, females weremore often body-dissatisfied than males [38], espe-cially regarding their own body weight [39]. Itcould be speculated that MtFs’ drive for thinness isa way to suppress masculinity and to correspond to

Figure 2 BUT GSI: comparisons among groups. Statistics—Values reported in the table are b values of Linear RegressionAnalyses, performed for between groups comparisons wih BUT GSI as dependent variable, and entering age, and BMI ascovariates, and groups as dummy variables (e.g., controls = 0; MtF w GRS = 1). *P < 0.05; **P < 0.01; ***P < 0.001. BUTGSI = Body Uneasiness Test Global Severity Index; MtF = male to female; FtM = female to male; GRS = genital reassign-ment surgery (w = performed; w/o = not performed); AN = anorexia nervosa; BN = bulimia nervosa; BED: binge eatingdisorder


J Sex Med 2013;10:1012–1023

Fig

ure

3B

ody

Une

asin

ess

Sub

scal

essc

ores

.B

UT

=B

ody

Une

asin

ess

Tes

tG

loba

lSev

erity

Inde

x;M

tF=

mal

eto

fem

ale;

FtM

=fe

mal

eto

mal

e;G

RS

=ge

nita

lrea

ssig

n-m

ent

surg

ery

(w=

perf

orm

ed;

w/o

=no

tpe

rfor

med

);A

N=

anor

exia

nerv

osa;

BN

=bu

limia

nerv

osa;

BE

D=

bing

eea

ting

diso

rder

1020 Bandini et al.

J Sex Med 2013;10:1012–1023

a female ideal of attractiveness [35,40–44], whereasFtM persons may be unwilling to lose weight, asbeing overweight makes breasts and hips looksmaller relative to abdominal size [17].

As far as compulsive body self-monitoring isconcerned, MtFs showed higher values than con-trols, as well as eating disorder subjects, whereasno differences were observed between FtMs andcontrols. Our findings confirmed Vocks et al.study [17], reporting that MtFs showed higherscores of compulsive body checking than FtMs andfemale controls, and are in line with previousobservations suggesting that such behaviors aremore represented in females than in males, both inclinical and general populations [42,43,45,46].

In summary, the findings related to BUT sub-scales suggested that GID subjects without anypreviously performed reassignment procedure dis-

played body uneasiness qualitatively similar toeating disorder subjects. Conversely, when the dif-ferent GID subsamples were taken into account,different dimensions of body discomfort seem tobe involved, depending on reassignment stage andgenotypic sex. In particular, it seems that GIDsubjects have a body attitude more similar to thegender congruent with their identity than withtheir genotypic sex.

As far as psychopathology is concerned, GIDsubjects did not show higher psychopathologicaldistress than controls, unlike eating disorderpatients. This finding was in line with most litera-ture data that have reported low levels of psycho-pathology in this clinical group [4–7] and seems tosupport the published preliminary recommenda-tions for DSM-V [47], which replaced the termGID with Gender Dysphoria, thus removing the

Figure 4 BUT GSI/SCL GSI ratio: comparisons among groups. Statistics—Values reported in the table are b values of LinearRegression Analyses, performed for between groups comparisons wih BUT GSI/SCI GSI ratio as dependent variable, andentering age, and BMI as covariates, and groups as dummy variables (e.g., controls = 0; MtF GRS = 1). *P < 0.05;**P < 0.01; ***P < 0.001. BUT = Body Uneasiness Test Global Severity Index; SCL-90 GSI = Sympton Checklist GlobalSeverity Index; MtF = male to female; FtM = female to male; GRS = genital reassignment surgery (w = performed; w/o = notperformed); AN = anorexia nervosa; BN = bulimia nervosa; BED = binge eating disorder


J Sex Med 2013;10:1012–1023

concept of “disorder.” However, MtFs w/oGRSshowed higher comorbidity than controls, inparticular adjustment disorder, in line with previ-ous findings [6]. As the difference with controlsvanishes in the group who performed GRS, psy-chopathology may result from the difficulty incoping with the social consequences of GID [6,10–12]. This hypothesis is also supported by the lowerprevalence of adjustment disorder in FtMs, as amale body has been considered more difficult topass off as a female one than the other way around[16], and male attributes seem to be more acceptedin females than female attributes in males [33].

To sum up, our findings show that a primarysource of suffering of GID subjects is their body,their levels of psychopathology are much lowerthan in eating disorder patients, and GRS mayremarkably reduce psychological distress.

Several limitations of the present study shouldbe acknowledged. The cross-sectional design ofthe study did not allow the detection of any causalrelationship between the variables taken intoaccount, and the specific effect of GRS on bodyuneasiness. Moreover, the main outcome measureswere based on self-reported questionnaires. In par-ticular, some items of BUT questionnaire mighthave a different meaning for GID and eating dis-order subjects. However, it should be noted thatthis specific instrument has already been used in asample of GID subjects [48]. Furthermore, itshould be taken into account that the relativelysmall sample size could lead to a reduced power indetecting significant differences that wereexpected (e.g., gender differences in body dissatis-faction). Finally, our findings refer to patientsseeking treatment and cannot be generalized tothe whole population with these syndromes.

In conclusion, GID and eating disorders arecharacterized by severe body uneasiness, and thebody is at the core of this distress in both condi-tions. Our findings suggest that in eating disorderpatients such uneasiness is primarily linked togeneral psychopathology, whereas in GID thisrelationship is lacking. These data suggest theneed for different, specific therapeutic strategiesfor body uneasiness in these two conditions.

Corresponding Author: Valdo Ricca, MD, PsychiatricUnit, Department of Neuropsychiatric Sciences, Flo-rence University School of Medicine, Largo Brambilla3, 50134 Firenze, Italy. Tel and Fax: +39-055-7947487;E-mail: [email protected]

Conflict of Interest: The authors have no conflicts toreport.

Statement of Authorship

Category 1(a) Conception and Design

Elisa Bandini; Alessandra Daphne Fisher; ValdoRicca; Mario Maggi

(b) Acquisition of DataElisa Bandini; Alessandra Daphne Fisher; ValdoRicca; Naika Ferruccio; Helen Casale; LorenzoLelli; Laura Benni; Maria Cristina Meriggiola;Davide Dettore

(c) Analysis and Interpretation of DataElisa Bandini; Alessandra Daphne Fisher; GiovanniCastellini; Carolina Lo Sauro; Valdo Ricca; MarioMaggi

Category 2(a) Drafting the Article

Elisa Bandini; Alessandra Daphne Fisher; ValdoRicca; Giovanni Castellini; Carolina Lo Sauro

(b) Revising It for Intellectual ContentElisa Bandini; Alessandra Daphne Fisher; ValdoRicca; Mario Maggi

Category 3(a) Final Approval of the Completed Article

Carlo Faravelli; Alessandra Daphne Fisher; ElisaBandini; Valdo Ricca; Mario Maggi

References

1 American Psychiatric Association. Diagnostic and statisticalmanual of mental disorders. 4th edition Text revision ed.Washington, DC: American Psychiatric Association; 2000.

2 Gooren L. The biology of human psychosexual differentiation.Horm Behav 2006;50:589–601.

3 Gooren L. Clinical practice. Care of transsexual persons.N Engl J Med 2011;364:1251–7.

4 Hoshiai M, Matsumoto Y, Sato T, Ohnishi M, Okabe N,Kishimoto Y, Terada S, Kuroda S. Psychiatric comorbidityamong patients with gender identity disorder. Psychiatry ClinNeurosci 2010;64:514–9.

5 Cole CM, O’Boyle M, Emory LE, Meyer WJ 3rd. Comorbid-ity of GID and other major psychiatric diagnoses. Arch SexBehav 1997;26:13–26.

6 Gómez-Gil E, Trilla A, Salamero M, Godás T, Valdés M.Sociodemographic, clinical, and psychiatric characteristics oftranssexuals from Spain. Arch Sex Behav 2009;38:378–92.

7 Haraldsen IR, Dahl AA. Symptom profiles of GID patients oftranssexual type compared to patients with personality disor-ders and healthy adults. Acta Psychiatr Scand 2000;102:276–81.

8 De Cuypere G, Janes C, Rubens R. Psychosocial functioningof transsexuals in Belgium. Acta Psychiatr Scand 1995;91:180–4.

9 Hepp U, Kraemer B, Schnyder U, Miller N, Delsignore A.Psychiatric comorbidity in gender identity disorder. J Psycho-som Res 2005;58:259–61.

10 Coleman E, WPATH Standards of Care Committee.Standards of care, 7th Version. 2012. Available at: http://www.wpath.org/documents/SOC%20V7%2003-17-12.pdf(accessed August 1, 2012).

1022 Bandini et al.

J Sex Med 2013;10:1012–1023

11 Nuttbrock L, Hwahng S, Bockting W, Rosenblum A, MasonM, Macri M, Becker J. Psychiatric impact of gender-relatedabuse across the life course of male-to-female transgenderpersons. J Sex Res 2010;47:12–23.

12 Matsumoto Y, Sato T, Ohnishi M, Kishimoto Y, Terada S,Kuroda S. Stress coping strategies of patients with genderidentity disorder. Psychiatry Clin Neurosci 2009;63:715–20.

13 Factor RJ, Rothblum ED. A study of transgender adults andtheir non-transgender siblings on demographic characteristics,social support, and experiences of violence. J LGBT HealthRes 2007;3:11–30.

14 Motmans J, Meier P, Ponnet K, T’Sjoen G. Female andmale transgender quality of life: Socioeconomic and medicaldifferences. J Sex Med 2012;9:743–50.

15 Wierckx K, Van Caenegem E, Elaut E, Dedecker D, Van dePeer F, Toye K, Weyers S, Hoebeke P, Monstrey S, DeCuypere G, T’Sjoen G. Quality of life and sexual healthafter sex reassignment surgery in transsexual men. J Sex Med2011;8:3379–88.

16 Pauly LB, Lindgren TW. Body image and gender identity.J Homosex 1976/77;2:133–4.

17 Vocks S, Stahn C, Loenser K, Legenbauer T. Eating and bodyimage disturbances in male-to-female and female-to-maletranssexuals. Arch Sex Behav 2009;38:364–77.

18 Johansson A, Sundbom E, Höjerback T, Bodlund O. A five-year follow-up study of Swedish adults with gender identitydisorder. Arch Sex Behav 2010;39:1429–37.

19 Fairburn CG. Eating disorders: The transdiagnostic viewand the cognitive behavioural theory; in Cognitive BehaviorTherapy and Eating Disorders. New York: Guilford Press;2008:7–22.

20 Ricca V, Castellini G, Lo Sauro C, Mannucci E, Ravaldi C,Rotella F, Faravelli C. Cognitive-behavioral therapy forthreshold and subthreshold anorexia nervosa: A three-yearfollow-up study. Psychother Psychosom 2010;79:238–48.

21 Zipfel S, Löwe B, Reas DL, Deter HC, Herzog W. Long-termprognosis in anorexia nervosa: Lessons from a 21-yearfollow-up study. Lancet 2000;355:721–2.

22 Castellini G, Lo Sauro C, Mannucci E, Ravaldi C, Rotella CM,Faravelli C, Ricca V. Diagnostic crossover and outcome pre-dictors in eating disorders according to DSM-IV and DSM-Vproposed criteria: A 6-year follow-up study. Psychosom Med2011;73:270–9.

23 Treasure J, Claudino AM, Zucker N. Eating disorders. Lancet2010;375:583–93.

24 Klump KL, Bulik CM, Kaye WH, Treasure J, Tyson E.Academy for eating disorders position paper: Eating disordersare serious mental illnesses. Int J Eat Disord 2009;42:97–103.

25 Dalle Grave R. Eating disorders: Progress and challenges. EurJ Intern Med 2011;22:153–60.

26 First MB, Spitzer RL, Gibbon M, Williams JBW. Structuredclinical interview for DSM-IV axis I disorders-patient edition(SCID-I/P version 2.0). New York: Biometrics ResearchDepartment, New York State Psychiatric Institute; 1995.

27 Derogatis LR. SCL-90-R; administration, scoring and proce-dure manual-II. Towson, MD: Clinical PsychometricResearch; 1992.

28 Prunas A, Sarno I, Preti E, Madeddu F, Perugini M. Psycho-metric proprieties of the Italian version of the SCL-90-R:A study on a large community sample. Eur Psychiatry2012;27:591–7.

29 Derogatis LR, Rickels K, Rock AF. The SCL-90 and theMMPI: A step in the validation of a new self- report scale. BrJ Psychiatry 1976;128:280–9.

30 Cuzzolaro M, Vetrone G, Marano G, Battacchi MW. BUT,Body Uneasiness Test: A new attitudinal body image scale.Psichiatria Infanz Adolesc 1999;66:417–28.

31 Cuzzolaro M, Vetrone G, Marano G, Garfinkel PE. TheBody Uneasiness Test (BUT): Development and validation ofa new body image assessment scale. Eat Weight Disord2006;11:1–13.

32 Marano G, Cuzzolaro M, Vetrone G, Garfinkel PE, TemperilliF, Spera G, Dalle Grave R, Calugi S, Marchesini G; QUO-VADIS Study Group. Validating the Body Uneasiness Test(BUT) in obese patients. Eat Weight Disord 2007;12:70–82.

33 Kraemer B, Delsignore A, Schnyder U, Hepp U. Body imageand transsexualism. Psychopathology 2008;41:96–10027.

34 Pauly IB, Lindgren TW. Body image and gender identity.J Homosex 1976–77;2:133–42.

35 Ålgars M, Alanko K, Santtila P, Sandnabba NK. Disorderedeating and gender identity disorder: A qualitative study. EatDisord 2012;20:300–11.

36 Kersting A, Reutemann M, Gast U, Ohrmann P, Suslow T,Michael N, Arolt V. Dissociative disorders and traumaticchildhood experiences in transsexuals. J Nerv Ment Dis2003;191:182–9.

37 Wolfradt U, Neumann K. Depersonalization, self-esteem andbody image in male-to-female transsexuals compared to maleand female controls. Arch Sex Behav 2001;30:301–10.

38 Garner MD. Body Image in Anorexia Nervosa. In: Cash TF,Pruzinsky T, eds. Body image, a handbook of theory, research,and clinical practice. 72 Spring Street, New York, NY: TheGuilford Press A Division of Guilford Publications, Inc;2002:10012.

39 Olivardia R. Mirror, mirror on the wall, who’s the largestof them all? The features and phenomenology of muscledysmorphia. Harv Rev Psychiatry 2001;9:254–9.

40 Hepp U, Milos G. Gender identity disorder and eatingdisorders. Int J Eat Disord 2002;32:473–8.

41 Sypeck MF, Gray JJ, Ahrens AH. No longer just a pretty face:Fashion magazines’ depictions of ideal female beauty from1959 to 1999. Int J Eat Disord 2004;36:342–7.

42 Haase AM, Mountford V, Waller G. Associations betweenbody checking and disordered eating behaviors in nonclinicalwomen. Int J Eat Disord 2011;44:465–8.

43 Alfano L, Hildebrandt T, Bannon K, Walker C, Walton KE.The impact of gender on the assessment of body checkingbehavior. Body Image 2011;8:20–5.

44 Hepp U, Milos G, Braun-Scharm H. Gender identity disorderand anorexia nervosa in male monozygotic twins. Int J EatDisord 2004;35:239–43.

45 Walker DC, Anderson DA, Hildebrandt T. Body checkingbehaviors in men. Body Image 2009;6:164–70.

46 Shafran R, Lee M, Payne E, Fairburn CG. An experimentalanalysis of body checking. Behav Res Ther 2007;45:113–21.

47 American Psychiatric Association. Sexual Dysfunction andGender Dysphoria, May, 2012. Available at: http://www.dsm5.org/ProposedRevision/Pages/proposedrevision.aspx (accessedAugust 1, 2012).

48 Bandini E, Fisher AD, Ricca V, Ristori J, Meriggiola MC,Jannini EA, Manieri C, Corona G, Monami M, FanniE, Galleni A, Forti G, Mannucci E, Maggi M. Childhoodmaltreatment in subjects with male-to-female gender identitydisorder. Int J Impot Res 2011;23:276–85.


J Sex Med 2013;10:1012–1023

gender identity disorder and eating disorders: similarities and differences in terms of body...

Documents